The intelligent generation of college schedule based on machine learning

In order to solve the problems of heavy workload and low efficiency of college resources scheduling. This article adopts reinforcement learning technology to schedule courses on a weekly week. The core of this method is the reward mechanism. By exploring the combination of hard conflict rewards and soft conflict rewards, we can find out the best action at which the average of the cumulative rewards reach maximum with each step, so as to generate the optimal schedule. In terms of computing power, the hybrid and parallel strategy is adopted. The process of learning and convergence is more stable and faster. The generated schedule performs well in many indicators, such as classroom utilization rate, the extent of the difference between the daily teaching time and the average. It has certain application value.